What Are Evidence Gaps in G3 History? | Reasoning Carefully When the Record Is Incomplete

CHAPTER 1 OF 16 · Locate the gap

1. An incomplete record is normal, but not equally incomplete everywhere

Back to contents

The past contained more actions, conversations and experiences than any archive could preserve. Some events were never written down. Some records were created and later lost. Some survive but are difficult to access, interpret or connect to a particular enquiry. Recognising this does not mean that all accounts of the past are equally uncertain. Different questions have different kinds and degrees of support.

A building’s opening date might be established through several independent records, while the private motives of one official remain unclear. A law’s wording may survive exactly, while evidence about its uneven enforcement is scattered. A military order might be documented, while the experiences of civilians affected by it are underrepresented. The gap must be described specifically. “We cannot know anything” is usually a much broader claim than the evidence permits.

Begin by naming the question, not by listing everything missing. An absent school register matters to an enquiry about attendance. It might matter less to an enquiry about the language of a public education speech. A missing record becomes an evidential gap through its relationship to a claim. This is why source relevance and uncertainty need to be considered together rather than as separate rituals.

Fictional teaching example: a town’s bridge plans and completion certificate survive, but daily crossing records do not. The surviving documents can establish design and recorded completion. They cannot directly establish how many people used the bridge each day. The absence of usage data limits one kind of conclusion, not every conclusion about the bridge. A careful answer preserves the distinction.

Students can make a simple map: secure findings, supported inferences and open questions. Keep claims in the appropriate area and move them when new evidence appears. This avoids two extremes: treating a gap as permission to imagine anything and treating it as proof that nothing can be said. Historical knowledge grows by matching the strength of each statement to the support actually available.

CHAPTER 2 OF 16 · Locate the gap

2. Ask how the record was created before asking what is missing

Back to contents

Sources were usually produced for contemporary purposes, not to answer a student’s later question. A tax register was designed to record taxable information; a diary reflected its writer’s interests; an official report followed an administrative task. The recording process determines what was likely to appear. Understanding that process helps explain why a source may be detailed about one matter and silent about another.

Ask who created the record, what information they needed, how they obtained it and what categories they used. An attendance register might count entries without explaining why people stayed away. A court record might preserve testimony through legal questions rather than a person’s uninterrupted account. A photograph might show visible conditions while leaving the events before and after the exposure unrecorded. These are properties to analyse, not automatic reasons for dismissal.

Fictional teaching example: a relief office records applications approved and rejected. The file may tell us how administrators classified requests. It may say much less about households that never applied, could not travel to the office or expected rejection. If the enquiry concerns total need, the record’s administrative boundary becomes important. The people outside the process are not necessarily absent from the population; they are absent from this recording system.

Do not assume that a category used by officials precisely captures the experience of everyone placed within it. Terms may reflect administrative priorities, assumptions or practical simplification. Read the definitions where available and compare them with other evidence. A historian can use the categories while explaining their limitations. The aim is to understand what the document measures, not to treat its labels as a transparent picture of society.

Once the creation process is clear, gaps become more intelligible. You can distinguish information the record never aimed to collect from information unexpectedly missing within its own purpose. That distinction affects how much weight silence should carry. It also suggests the next source to seek: a different institution, a personal account, material evidence or another record that captures people outside the first system.

CHAPTER 3 OF 16 · Locate the gap

3. Separate non-creation, loss, selection and access

Back to contents

A missing source can be missing in several ways. It may never have existed. It may have been destroyed accidentally or deliberately. An archive may have selected only part of a collection. A relevant record may survive but remain uncatalogued, restricted, untranslated or outside the material available for the task. These possibilities should not be treated as interchangeable.

The difference matters because each suggests a different limit. If no minutes were ever taken, the discussion cannot be recovered through a minute book. If minutes survive but are unavailable, the present enquiry has an access limit rather than proof that the information never existed. If a file is incomplete, the missing section may change how the surviving pages should be interpreted. Be precise about what is known regarding the absence itself.

Fictional teaching example: a council archive contains meeting books for every year except one. Without further evidence, you can say that the year’s book is absent from the collection consulted. You cannot confidently say it was destroyed to hide wrongdoing. That is one possible explanation among many, and a serious allegation would require evidence about destruction, custody or motive. The gap is real; its cause remains uncertain.

Avoid making archival survival into an intention by default. Some records endure because institutions had storage resources, because copies were distributed or because later collectors valued them. Others disappear through routine disposal, damage or neglect. Deliberate suppression can occur, but it should be established rather than assumed. A careful historian asks about the chain of preservation and recognises where that chain is unknown.

For student writing, a short phrase often suffices: “The supplied sources do not establish…” or “The surviving file consulted here lacks…” These formulations keep the claim within the evidence available. They are more accurate than announcing that “there is no evidence anywhere.” Unless you have conducted a comprehensive search, distinguish the limits of the pack or enquiry from the total state of historical knowledge.

CHAPTER 4 OF 16 · Use absence carefully

4. Silence matters only when something was reasonably expected

Back to contents

An argument from silence draws meaning from a missing mention. It can be useful, but only when there is a reason to expect the source to record the matter. A complete ledger’s failure to list an expected payment may raise a significant question. A brief ceremonial speech’s failure to discuss that payment may mean very little. The same absence has different weight under different recording purposes.

Test four things: relevance to the record’s function, completeness of survival, the creator’s opportunity to know and the likely reasons for mentioning or omitting it. You do not need to recite these as a formula in every answer. Use them to decide whether the silence should carry an argument or merely prompt further investigation. The more uncertain the expectation, the weaker the inference from absence.

Fictional teaching example: an inspection form requires every damaged classroom window to be listed. A complete signed form records no damage, while a later memoir describes broken windows. The silence on the form is relevant because the information was specifically required. Even then, the inspector may have overlooked a room, misunderstood the instruction or recorded the condition before damage occurred. The contradiction needs testing rather than an automatic verdict.

Contrast this with a school anniversary brochure that says nothing about broken windows. A celebratory publication had no comparable obligation to record maintenance problems. Its silence may reveal selection if the enquiry concerns public image, but cannot securely establish the building’s condition. The question determines the use, and the source’s function determines what absence can reasonably suggest.

A proportionate sentence might say that the omission weakens a particular claim because a complete inspection should have recorded the damage, while noting uncertainty about timing or thoroughness. This is a reasoned use of silence. “It is not mentioned, so it did not happen” skips every necessary test. Historical caution here is not pedantry; it prevents the record’s limits from becoming false certainty.

CHAPTER 5 OF 16 · Use absence carefully

5. Missing voices require investigation, not invented testimony

Back to contents

Some groups are less visible in surviving records because they had fewer opportunities to create documents, because their materials were not preserved or because institutions recorded them mainly through outside categories. This can distort the apparent balance of perspectives. A large volume of official correspondence does not mean officials were the only people with experiences or opinions worth investigating.

Start by identifying whose direct testimony is absent and why that matters to the enquiry. Then look for sources through which their actions or experiences may still appear. Petitions, court testimony, wage records, photographs, objects, housing evidence and oral histories can provide different forms of access. None should be assumed to speak transparently for an entire group. Each requires attention to creation, mediation and scope.

Fictional teaching example: a housing project’s administrative files describe relocation targets, but few residents’ letters survive. A student might examine compensation claims, transport changes and household records to investigate likely burdens. These can support conclusions about costs or access. They cannot justify invented quotations about residents’ feelings. Material conditions and subjective experience overlap, but one does not automatically establish every detail of the other.

Be careful with a sympathetic generalisation. “The residents must have felt…” can sound humane while replacing missing people with the writer’s own assumptions. Better wording identifies what is supported: longer journeys, recorded complaints, changes in household arrangements or stated concerns in the testimony that survives. Where feeling is inferred, label the inference and explain its basis. Respect includes not claiming to know what the evidence does not tell us.

A gap in direct testimony should remain visible in the conclusion. You can say that the administrative account is incomplete for evaluating residents’ experiences and identify the kinds of evidence needed. This does not make the official material useless. It tells the reader what it can establish and where another perspective is needed. A fuller history often begins with noticing that the archive’s loudest voice is not the whole conversation.

CHAPTER 6 OF 16 · Use absence carefully

6. Corroboration reduces some gaps but cannot repair every one

Back to contents

Corroboration means checking a claim against other evidence. It can strengthen confidence, reveal disagreement or identify a narrower version of the claim that survives testing. However, several sources are not automatically several independent confirmations. Newspapers may repeat the same official statement. Later accounts may copy an earlier memoir. Before counting agreement, ask whether the sources could share an origin.

Fictional teaching example: three articles report that a bridge carried ten thousand passengers on its first day. If all three quote the same press release, they provide evidence of the claim’s circulation, not three separate measurements. A toll register or independently conducted count would test the number differently. The relationship between sources matters more than the simple number of items that agree.

Corroboration also needs a common claim. One source recording high attendance and another describing public enthusiasm are related but not identical. Attendance may have several causes; enthusiasm may concern only part of the crowd. Explain which detail is confirmed and which remains uncertain. Otherwise, a vague phrase such as “the sources support each other” can conceal a gap that neither source actually closes.

Disagreement can be productive. Different dates, locations or definitions may explain apparently conflicting figures. A contemporary report may count registered participants while a later account estimates everyone present. Once definitions are aligned, the conflict may narrow or disappear. If it remains, investigate access, purpose and method rather than automatically preferring the source that fits your initial argument.

A strong conclusion can distinguish a well-supported finding from its unresolved explanation. Several independent sources may establish that a protest occurred, while motives remain contested. Corroboration has secured the event without settling every cause. Recognising that uneven progress keeps historical knowledge realistic: some questions can be answered confidently while related questions still require caution.

CHAPTER 7 OF 16 · Evaluate explanations

7. Distinguish fact, inference and hypothesis

Back to contents

A documented fact is a claim supported by evidence appropriate to the enquiry, though it may still be revised if better information emerges. An inference is a reasoned conclusion drawn from details rather than stated directly. A hypothesis is an explanation proposed for testing. These distinctions are practical ways to control uncertainty. They do not imply that facts require no interpretation or that every inference is equally weak.

Fictional teaching example: minutes record that a council postponed a vote after receiving a cost estimate. The sequence is documented. It is reasonable to infer that cost may have influenced the postponement, especially if discussion focused on funding. A hypothesis that the delay was arranged to help a private company requires additional evidence. The further the explanation moves beyond what is recorded, the more clearly its supporting steps must be shown.

Use language that reflects the status of the claim. “The minutes record” identifies a statement in a source. “This suggests” introduces an inference. “One possible explanation is” opens a hypothesis. These phrases should not become empty verbal habits. Follow them with evidence and a reason. Saying “possibly” does not make an unsupported story responsible if no basis for the possibility is provided.

An important check is whether a hypothesis could be tested. What evidence would distinguish it from another explanation? Correspondence about costs might support financial concern; records of private negotiation might support another motive. If every possible source could be interpreted as confirming the theory, it may have become insulated from evidence. Historical reasoning improves when it states what would count against an explanation as well as what supports it.

In student writing, do not give a speculative explanation the same weight as an established finding merely because both fit the narrative. Organise the paragraph so that the reader sees the difference. A clear sentence can say that the timing is consistent with financial concern but the sources do not establish whether it was the decisive reason. This preserves a useful inference without pretending the gap has disappeared.

CHAPTER 8 OF 16 · Evaluate explanations

8. Test rival explanations instead of choosing the most vivid story

Back to contents

An incomplete record often permits more than one explanation. The task is not to list every imaginable scenario. It is to compare plausible alternatives that fit the known circumstances and ask which best accounts for the evidence. A vivid story may be memorable without being well supported. Explanatory strength depends on evidence, mechanism, consistency and the number of unsupported assumptions required.

Fictional teaching example: attendance at a training centre falls. Possible explanations include inconvenient hours, fees, reduced demand or a change in record keeping. These are not equally likely simply because they can be named. A timetable change, payment records, local employment figures and administrative instructions could help distinguish them. The student should connect each alternative to evidence rather than treat speculation as a balanced answer.

Look for an expected pattern. If higher fees discouraged attendance, did the decline occur after the increase and particularly among those most affected? If working hours were the main problem, do complaints or attendance differences fit that explanation? These tests do not always produce certainty, but they make the reasoning more disciplined. An explanation that predicts no observable difference is harder to evaluate.

Avoid changing the evidence to rescue a preferred theory. If records contradict your initial view, revise the view or explain the contradiction with support. Do not invent hidden motives every time a source disagrees. A good hypothesis remains open to rejection. The willingness to adjust an explanation is a strength of historical enquiry, not a failure to defend an argument.

A proportionate conclusion might state that the fee explanation is better supported than the timetable explanation because the timing and affected groups align, while noting that the available records cannot exclude an additional influence from local employment. This is more useful than either claiming one complete cause or ending with “anything could have happened.” You have compared alternatives and explained the remaining uncertainty.

CHAPTER 9 OF 16 · Evaluate explanations

9. Material and numerical evidence can illuminate different gaps

Back to contents

Written testimony is not the only route into the past. Buildings, objects, maps, photographs and numerical records can reveal patterns that individual accounts leave unclear. They may help establish where people lived, what goods circulated or how an institution operated. Their value depends on the question and on careful interpretation. An object does not arrive with a complete explanation of its use or meaning.

Fictional teaching example: household accounts are missing, but housing plans and water connection records survive. These might help investigate access to infrastructure. They do not automatically establish how residents felt about their homes or whether every connection worked reliably. The sources close part of the material gap while leaving other questions open. A good answer states both the contribution and the boundary.

Numbers require definitions. Does a register count households, individuals, visits or applications? Are the categories consistent across years? Are missing entries treated as zero, unknown or excluded? A rising total could reflect expanded coverage rather than actual growth. Before drawing a trend, examine whether the measurement process changed. Otherwise, the attempt to fill a gap may create a new error.

Maps also need attention to purpose and date. A planned road is not necessarily a completed road. An administrative boundary is not necessarily a boundary of everyday movement or identity. A photograph records a selected view at a particular moment. Each source can provide valuable evidence without being a transparent window onto everything outside its frame.

When combining source types, assign each a job. A map establishes distance; an account records travel cost; a letter describes a stated complaint. Their convergence may support an explanation of access difficulties. Do not make one source prove what another merely suggests. The strongest synthesis preserves the differences between evidence types while showing how they contribute to a common enquiry.

CHAPTER 10 OF 16 · Reach a conclusion

10. Oral histories and later memories need temporal care

Back to contents

A later recollection can preserve experiences that were not recorded at the time. It can also reflect memory, later knowledge, changing interpretation and the circumstances of the interview. These features do not make oral history worthless. They determine how it should be used. Distinguish evidence about a past event from evidence about how that event was later remembered and understood.

Fictional teaching example: a retired worker recalls a factory closure many years afterwards. Their account may provide details of daily routines, family effects and personal meaning absent from company files. Exact dates or reported conversations may need checking against other records. If the worker interprets the closure through later developments, identify that retrospective perspective rather than treating it as knowledge held at the moment.

Ask how the testimony was produced. Who asked the questions? Were prompts leading? Was the account edited, translated or excerpted? Was the speaker discussing a public controversy or a private memory? These questions help evaluate particular claims without reducing the person to a generic label of unreliable memory. Administrative documents also have production processes; oral testimony deserves analysis, not automatic suspicion.

Do not assume that one remembered experience represents a whole workforce or community. It may illuminate a significant variation precisely because it differs from the dominant account. Compare it with other testimony and material where possible. If only one account survives, explain its value for that person’s experience and the limits on broader generalisation.

A careful sentence might say that the recollection reveals how the worker later understood the closure’s impact, while contemporary payroll records support the timing of job loss. These sources do different work. Together they can enrich the account without forcing memory to supply exact administrative data or forcing a payroll list to describe emotional experience. Recognising distinct evidential roles makes incomplete histories more informative.

CHAPTER 11 OF 16 · Reach a conclusion

11. Write a conclusion that is useful without pretending completeness

Back to contents

A conclusion should answer the question as far as the evidence permits. It should not become a catalogue of everything historians might someday investigate. State the strongest finding, explain the main reason for it and identify the specific uncertainty that affects its scope or confidence. The gap belongs in the conclusion only when it matters to the judgement being made.

Fictional teaching example: evidence shows that a relief programme distributed food in three districts, while recipient lists for two others are missing. You can conclude that distribution occurred in the documented districts. You cannot infer zero distribution in the others merely from missing lists. If the question asks about overall coverage, the missing records limit a town-wide estimate. If it asks about procedures in one documented district, the same gap may be less central.

Use bounded wording rather than blanket disclaimers. “The available records establish X but do not settle Y” is often clearer than “All sources are biased, so we cannot know.” The latter abandons the distinctions built during the enquiry. Bias, incompleteness and interpretation are reasons to evaluate evidence carefully, not reasons to treat every account as equally unsupported.

Avoid a false compromise. If one explanation has substantially stronger evidence, say so even if certainty is impossible. Historical judgement does not require absolute proof before preferring a conclusion. Explain why it is better supported and which unresolved point could change it. This gives the reader a decision they can inspect rather than a vague refusal to decide.

A helpful final check is to compare the conclusion’s certainty with the body of the answer. Have you moved from “suggests” in the analysis to “proves” at the end without new support? Have you suddenly become more doubtful than the evidence warrants? Consistency matters in both directions. The conclusion should reflect the reasoning already demonstrated, not intensify or dissolve it for rhetorical effect.

CHAPTER 12 OF 16 · Reach a conclusion

12. A complete fictional incomplete-record workshop

Back to contents

Here is a fictional case, not a historical account. A town introduces a subsidised ferry. The surviving launch speech promises affordable travel. A register records ticket sales for the first two months. A newspaper carries complaints about cancellations. The operator’s maintenance files are missing. The enquiry asks whether the ferry provided a dependable service during its first year. Notice that both the time coverage and the measure of dependability need attention.

Begin with what is established. The speech records a stated ambition about affordability, not dependable operation. Ticket sales establish recorded use during two months, assuming the register is sound. Complaints provide evidence of reported cancellations, but may not reveal the total number of sailings or the frequency of failure. The missing maintenance files limit investigation of mechanical causes, although other sources might record operations.

Now identify what the pack cannot support. It cannot calculate an annual cancellation rate from a few complaints and a short sales record. It cannot prove that all cancellations were mechanical. It cannot show that high sales mean reliable service; demand may remain strong when alternatives are poor. Each rejected inference protects the answer from using one source to fill a gap it cannot actually close.

A provisional conclusion could say that the complaints raise a credible concern about dependability, but the available pack does not establish performance across the whole year. This is not an empty answer. It separates a supported concern from an unsupported overall rate. The next evidence sought might include sailing logs, harbour records, passenger accounts from different months or payment records that distinguish scheduled and completed journeys.

For practice, change the question to “How was the ferry promoted?” The speech becomes central and the maintenance gap matters less. Change it to “Why were sailings cancelled?” The missing files become more consequential, but weather records or harbour notices might still help. The exercise demonstrates that a gap is not simply a missing object. It is a limitation on a particular inference within a particular enquiry.

CHAPTER 13 OF 16 · Continue enquiry

13. Common errors and the repair each one needs

Back to contents

The first error is filling silence with certainty: “There are no complaints in the pack, so everyone was satisfied.” Repair it by asking whether complaints were sought, recorded and preserved. The second is conspiracy by default: “The file is missing, so someone hid it.” Repair it by distinguishing the absence from its unestablished cause. A serious claim about deliberate concealment requires evidence of its own.

The third error is helpless scepticism: “The record is incomplete, so no conclusion is possible.” Repair it by separating well-supported findings from unresolved questions. The fourth is decorative uncertainty: adding “perhaps” before an elaborate invented motive. Repair it by identifying the evidence that makes the possibility plausible and the evidence that could test it. Cautious wording cannot replace a reasoning chain.

The fifth error is source counting. Three dependent reports are treated as three independent confirmations. Repair it by tracing origins and comparing methods. The sixth is scope expansion. One person’s testimony is made to represent an entire population. Repair it by naming the person or group actually covered and stating what would be needed to generalise further. These corrections are specific, learnable operations.

Fictional teaching repair: replace “The missing register proves the scheme failed” with “The missing register prevents a complete estimate of participation, although the surviving correspondence documents difficulties in two districts.” The revised sentence preserves useful evidence while removing an invalid conclusion. It also tells the reader what kind of uncertainty remains: coverage, not necessarily the existence of every difficulty.

When revising an essay, label each uncertain claim with its main issue: missing source, narrow coverage, ambiguous wording, conflicting testimony or uncertain causation. Different issues need different remedies. A date check will not solve a representativeness problem; another quotation from the same author may not solve dependence. Diagnosing the gap accurately makes revision more productive than adding a generic paragraph about limitations.

CHAPTER 14 OF 16 · Continue enquiry

14. A practical routine for the next source-based task

Back to contents

Start with a one-sentence enquiry and a list of claims needed to answer it. Beside each claim, note the strongest available evidence and one important limit. Keep the list small enough to use. The purpose is not to create an elaborate research system for every short question, but to make the relationship between claim, support and uncertainty visible before drafting.

During writing, distinguish source statements from your conclusions. Attribute claims, explain inferences and avoid inventing missing dialogue or motives. Where sources disagree, compare scope and method before deciding that one must be entirely wrong. Where evidence is absent, ask whether the absence matters under the source’s expected function. These checks keep the paragraph focused on the actual enquiry.

After writing, highlight certainty words: proves, all, none, only, definitely and impossible. Check whether the evidence supports the exclusions each word makes. Then highlight vague disclaimers such as “we can never know.” Check whether they throw away findings already established. Good historical writing controls overconfidence and unnecessary scepticism together. It does not assume that caution always means saying less.

A parent can help by asking, “Which part do you know, which part are you inferring, and which part still needs evidence?” The student’s ability to explain those distinctions is a useful sign of progress. There is no need to turn every answer into a debate about complete certainty. The aim is a reasoned conclusion at the right level, following the school’s current task requirements.

The lasting lesson is straightforward: an incomplete record demands visible reasoning. Establish what you can, test plausible explanations, name the consequential gaps and stop before invention begins. History remains meaningful because evidence can support strong conclusions even when it cannot answer every question. Learning where to place that boundary is a core intellectual habit, not a weakness in the subject.

CHAPTER 15 OF 16 · Continue enquiry

15. Turn an unanswered question into a focused next enquiry

Back to contents

A useful research question names the missing relationship rather than asking vaguely for more information. “Find more sources about the ferry” is broad. “Find monthly records that distinguish scheduled sailings from completed sailings” targets the dependability gap. The second instruction tells you what a successful source must contain and how it would change the argument. This makes further enquiry purposeful rather than endless.

Prioritise gaps by their effect on the conclusion. Some missing details would add colour without changing the answer. Others determine whether the central claim holds. If the question concerns annual reliability, operating records across the year matter more than another launch photograph. If it concerns public image, the photograph may become important. Research priorities follow the enquiry just as evidence selection does.

Also ask whether the gap can reasonably be closed within the task. A classroom source pack may be deliberately limited. You are usually expected to reason from the supplied material and relevant knowledge, not conduct an archive search during an assessment. State the limitation briefly and continue with the supported answer. Do not let an imagined research project displace the question you were actually asked.

In independent study, keep a short question log. Record the uncertainty, the kind of evidence needed and what finding would support or weaken the current explanation. When a new source appears, test it against that log. Does it close the gap, narrow it or merely repeat an existing claim? This routine helps students recognise that obtaining another document is not the same as obtaining the information needed.

An unanswered question can therefore be a productive result. It identifies the boundary of current knowledge and gives the next enquiry a clear direction. The important thing is to keep that question honest: neither disguise it as a fact nor enlarge it into doubt about everything already established.

CHAPTER 16 OF 16 · Continue enquiry

16. Keep uncertainty specific during class discussion

Back to contents

When a classmate offers a different conclusion, ask which part of the record supports it and which gap they think matters most. You may discover that you agree on the facts but weigh a limitation differently. That is a more productive disagreement than deciding that one person is simply confident and the other cautious. Historical discussion improves when each claim has an identifiable evidential basis.

Use a three-part response: “We agree that…”, “The uncertainty concerns…” and “The evidence that would help is…”. These prompts organise the discussion without forcing a compromise. One explanation may still be better supported. The point is to locate the disagreement accurately so that it can be examined rather than repeated.

If the missing evidence is unavailable, return to the task and state the strongest conclusion both the evidence and its limitations permit. A classroom discussion does not need to end with total agreement to be successful. It can end with a clear understanding of why one claim is stronger, why another remains possible and why a third goes beyond what the sources can support.